**The data files and programs are written with Stata 9.0 SE

**************************************
The data set definitions are:
******WTI DATA ***********
wti_clean.dta contains WTI daily price changes. The variables in the data set are:
wti=WTI price level
price_ch=100*log(p(t)/p(t-1))
******Retail Gas price DATA ***********
retail_gas_daily_clean.dta contains retail gasoline price changes. The variables in the data set are:
retail_gas=Retail gas price level
price_ch=100*log(p(t)/p(t-1))
******Macroeconomic Announcement Data ***********
mms_all_final2.dta contains macroeconomic news surprise data. The variables in the data set are:
year=year of the release of the announcement
month=month of the release of the announcement
day=day of the release of the announcement
actX=actual macroeconomic announcement X
foreX=forecast of macroeconomic announcement X
surX=actX-foreX
sursX=(actX-foreX)/stdev(actX-foreX)
The X stands for different macroeconomic announcements numbered as follows:
*GDP Advanced (26), GDP Preliminary (27), GDP Final (28), Unemployment Rate (1), Nonfarm Payroll (7),
 Retail Sales (9), Industrial Production (11), Capacity Utilization (10), Personal Income (19),
 Consumer Credit (21), Personal Consumption Expenditures (17), New Home Sales (18), 
 Durable Goods Orders (3), Construnction Spending (13), Factory Orders (15), Business Inventories (12),
 Government Budget Deficit (20), Trade Balance (6), PPI (8), Core PPI (34), CPI (2), Core CPI (33)
 Consumer Confidence (U. Mich.) (P 71, F 72),  Consumer Confidence Index (Board) (14), NAPM Index (16),
 Housing Starts (4), Index of Leading Indicators (5), Target Rate Surprises (32), Initial Claims (25)
************************************


**************PROGRAMS***************
regress_daily_one_at_a_time_wti.do
This stata program reproduces table 2a in the paper titled: "Daily WTI Crude Oil Prices: Individual Regressions for 1983-2008"
regress_all_announcements_on_wti.do
This stata program reproduces table 2b in the paper titled: "Daily WTI Crude Oil Prices: Joint Regressions for 1983-2008"
It also contains the joint significance tests shown in Table 5 titled "p-values of Joint Significance Tests for 1983-2008"
regress_daily_one_at_a_time_retail.do
This stata program reproduces table 3a in the paper titled: "Daily US Gasoline Prices: Individual Regressions for 2003-2008"
regress_all_announcements_on_retail.do
This stata program reproduces table 3b in the paper titled: "Daily US Gasoline Prices: Joint Regressions for 2003-2008"
It also contains the joint significance tests shown in Table 5 titled "p-values of Joint Significance Tests for 1983-2008"
regress_all_announcements_on_wti_monthly.do
This stata program is used to compute the joint regression p-values in table 4 in the paper titled: "Monthly WTI Crude Oil Prices: Regressions for 1983-2008" 
It also contains the joint significance tests shown in Table 5 titled "p-values of Joint Significance Tests for 1983-2008"
regress_monthly_one_at_a_time_wti.do
This stata program is used to compute the individual regression p-values in table 4 in the paper titled: "Monthly WTI Crude Oil Prices: Regressions for 1983-2008"
regress_diff_horizon_one_at_time_retail_continuous2.do and regress_diff_horizon_one_at_time_wti_continuous3.do 
These programs are used to create Figure 1
regress_monthly_one_at_a_time_wti.do
This stata program is used to compute the individual regression p-values in table 4 in the paper titled: "Monthly WTI Crude Oil Prices: Regressions for 1983-2008"
The programs record3.do and record.chris.one_tail.do are used by other programs to output the estimates into a table format


**************************************************Matlab programs
The bootstrap code is written in Matlab. The files

	table2a.m
	table2b.m
	table3a.m
	table3b.m
	table4a.m (individual)
	table4b.m (joint)

generate the respective tables in the paper. The last line 
of these files evaluated at the t-statistic produces the data-
mining robust p-value.

The file
	figure1.m

produces Figure 1 in the paper. The raw data was produced in Stata.
